Trying to find anyother limits on PE PW in 8d as cannot get over set amounts per inj constant setting.

I have tried setting PE AFR to 9.5 and VEs to 100 with no luck.

27 inj \ 27 constant = 11.40 ms
30 inj \ 30 constant = 10.27 ms
30 inj \ 29 constant = 10.90 ms
30 inj \ 27 constant = 11.40 ms
37 inj \ 37-34 constant = 8.20-8.70 (memory)

Have studied PE WOT tuning thread and tried Traxe's VE program to check AFRs.

If anyone could point me to more info on this flatlining PW thing, would greatly appreciate.

Thanks guys.

 

You lost me. I don't get what your asking.

 

Hi Grumpy,
I can not get inj pw to increase over 11.40 ms with current 30lb injectors set at 27 inj constant. Would only go to 10.27 ms at 30 lb inj constant. This seems to happen in PE anywhere between 2000 rpm and up to 5000 shift point.

Have been trying to richen to 900-1000 mv as had with TBI 7747 back in 97-98 to see if helped KS.

Have tried 27, 30, and 37 lb injectors at like inj constant settings but mv stay between 750-825 and KS.

This is a 502-502 crate engine with tuned port in my Suburban.

Would the Fuel Limit Factor vrs Coolant table @ L8563 that controls AE and DE be effecting PE?

Don't know how else to explain. Have searched through ANHT hac several times to no avail.

PS: Have seen PW flash to 11.9-12.5 ms while scanning in PE for split second but drops to 11.4 and flatlines.
Thought maybe need different terminology in search function.

 

A 747 fires the injectors alternatively per every DRP. At 11 msec that would be similiar to a 730 with a 22 injector PW, and that would mean the injector being static at like ~3,000 RPM.

What injectors are you trying to run?.

The table starting at 563 only looks to be AE related. The one at 671 should allow you to get all the PW you chould need.

If your trying to run 2 P+H TBI injectors with a 730, you're trying to force them to hand twice their design load. The 749 wired up as used in the Sunbirds can handle that injector load.

 

Grunpy,
Tried to give full info but lost post second time.

Canot find 671 but will continue to look.

7730 with 8d and (8) 30 lb Lucas set at 27 constant to get 11.40 PW.

Tuned port 502 big block.

Thanks.

PS: Was comparing 7747 for 02 mvs only. Did not expect same pw as 7747 pulses differently.

 

Bob,

This is definitely a problem and I don't know the answer yet. I have a buddy (Hi Rick!) who has a 427 8D'd thirdgen with 42lb/hr injectors. The Pulse Width flatlines at 6.85ms. It's insane. I change the PE vs. RPM or PE vs. Coolant tables and the REQUESTED AFR changes but the Pulse Width remains constant and the same. You just can't get it to go above 6.85ms. We tried a fresh AUJP, a new ECM, etc. 6.85ms. We ALSO tried changing the VE tables by increasing the VE. Still 6.85ms! Crazy. Something is capping the PW calculation. Note that it is NOT capping the requested AFR calculation. It has to be in the code after that. I looked at the ANHT hac for hours trying to track this one and couldn't find it. It might end up just being an AUJP thing (just don't know). One day I'll get back to it ... just not as important right now because that 6.85ms is plenty of fuel so we don't really NEED to go higher.

Tim

 

Thanks Tim,

Thought it was something like that but glad someone else has seen it.

Guess will just have to fudge inj constant down untill can get PWs up high enough to hit 950-1000 mv in PE.

Just experimenting with fuel and spark combos and couldn't get it with higher FP and current fuel pump.

Thanks again Tim and Grumpy.

 

Guess it's just another reason for running the 58 code, or putting together an ecm bench and tracking thru the 8D code.

 

Bob,

FWIW, my only suggestion would be that it might have something to do with the "Fuel Limit Factor vs. Coolant" table. This table is at 563. Check the actual code beginning at LE94C. I cannot say for sure that this is it. I just don't know. But, this is where I ended up after tracking through everything that I could think of

Tim

 

Thanks Tim,

I asked Grumpy about that table and he indicated that it infact did only effect AE and DE (no help with PE).
May try changing 56C and up settings to match lower coolant settings to see results.

May also try increasing cyl cc instead of lowering inj constant so can keep correct inj constant per inj size. Don't know which method messes up other calcs in mask the least.

 

Bob,

The cylinder size had an affect just like you changing the injector costant. SOMEWHERE there is a limit based on the total amount of fuel. Changing the injector constant or the cylinder size lies to the computer so it uses a different max BPW (but probably still the very same total amount of fuel). I don't have the resources to play with this anymore (no car or 730 system available). However, it would definitely be cool for someone to keep focused on it and figure it out. Rarely does a pure unknown with the 730 come out of nowhere at this point in the game. This is a constant or table (even if it is hardcoded) that none of us have specifically stated yet. I only joined in this topic because finally somebody else is experiencing the same thing.

Tim

 

I've looked through the AXCN code and didn't see anything either. I wonder if maybe this is an AUJP only item, such as the 800 RPM idle limiter.

Maybe try out one of the Y Body bins and see if it does the same thing. Can just cut & paste the cal data as all that is in the same place.

RBob.

 

I just spent some time on the ecm bench, with a AUJP. Disabled the VATS, and a few other codes that were setting, set the PE enrichment to max, and got a 12 msec PW at ~6K RPM, and it was ~14 at 3K RPM.

Using TTS Datamaster.

I just set the VE and VE ext to 100%s (in addition to max'ing out the PE AFR) and from 1,600 RPM to 6,375 had a PW of 13.96 msec..

 

Use a large cylinder volume and a large injector size (as if you were running a big block). I would recommend an injector size of 40+ along with a big cylinder volume (427ci +)

Tim

 

I don't know if this is it, or if my question has been explained or not.But here it is.
Look at the listed address in the anht and the next page. I din't understand what was going on here so I asked.
8563?
I got some more ref material on assembly, but haven't had time to go back and research this more after I've read some of it. I did find a max # that was like 50 or 500, but that's not it. The address I listed is the start of the 255 g/s air stuff which was posted earlier by SMasterson(sp) as not a limit.

Have you tried actually doing the PW calc yourself based off your logs? Might be able to find the problem the hard way.
That's what I was doing just as a learning experience. My ads file doesn't do the PW calc right or I could test your bin for you.

 

Have looked back through a few years of Diacom logs (reading with Datamaster) on this engine and see same actons on ANHT based bins, TPIS, AS&M, S&Perf, and Superchips based bins have purchased. The ANHT, AUJP, and S-AUJP currently playing with are only ones that I know the what the base prom is. As I tuned each base bin, I tuned VEs fairly close. Of course had to retune VEs for each inj size and inj constant setting. Tried to leave 100 KPA VE tables alone to not lower PE fueling (ie PW).

This Suburban have been working on since approx 97 and is second 502 engine in it. Only play with during summers in Maine.
Winters are spent working on my 2 S-10s with 7727 and 7749 (8d and 58).

Looked back at some of the 420SBC TPI logs and am seeing same PW flatlining but not as bad.

Lowering WOT timing to 20-25 total seems to have stopped KS but now have cracked headers and feels sluggish.

My Thorly hdrs are no longer available so looks like tuning will stop until can locate something for exhaust as may be messing up my VE adjusting. BLMs are acting strange.

Thanks all for the help.

 

Any new information on this yet? Played with the same car this weekend and observed the same things. Basically, no control over WOT PE Base Pulse Width. The lower RPMs were running lean and we tried to richen them up. Even went so far as to add 10% VE to the lower RPMs and the AFRs stayed 100% the same. There is just no WOT PE fuel control. No matter what you change in the PE or VE tables the ECM just does whatever it wants to do. If I had an ECM bench I could lick it and find out very quickly. However, no ECM bench here.

Tim

 

Tim,
Never did find a solution so just used 27 constant for 30 inj but still not rich enough.
Back to Fla. now and left 502 in Maine. Will try 37 inj with 30 constant next summer. Back to work on my 420 SBC with 8d and 383 with 58d. 58d. doesn,t seem to limit pw as seeing 11.4 with 37 inj but of course constant does not apply in 58d.

Am selling 383 blower Blazer as been 4 years and can not straighten out. Was going to return to 8d. but got fed up.

Now will work on 420 engine some more. May try fudging cyl size instead of inj constant to see if doesn.t mess up all BLMs as much.

Wish someone knew about this pw limit deal on 8d. May still play with Fuel Limit factor vrs coolant to varify doesn't do anything (L8563).

Bob

 

here's a thought..

what if the limit is not intentional, per se; ... but is instead some variable maxing out in memory? the MAF 255 g/s limit is well known, but perhaps the calculation which calculates the total amount of fuel to be injected at some point has a value which maxes at 255.

i don't know the exact equation to calculate the PW, but it seems that the max PW achievable varies based on the cylinder size/injector size....so lets say the equation is something like this

(cyl size * VE * BLM/128 * INT/128) /inj const

if at some point the value in the parentheses () exceeded 255, and simply gets pegged there (like a MAF at 255g/s)... the actual pulsewidth achievable still varies because you divide 255 by the injector constant..?

if there were a way to watch the memory spots on an ecm bench, you could start at idle rpm and slowly ramp up....and as soon as the PW flatlined, you could find the memory spot which also quit increasing/got stuck, and then trace back where it's used to. just a thought - i have no idea how practical/real any of the above is, i don't have an ecm bench yet... nor have i hit this problem with my stock 350.

 

91Z,

This is exactly what I think is going on. In fact, this is how I originally moved forward on this. I was looking for exactly what you said. Initially, I found a location which limits the total airflow to 255! I thought this was it. But then RBob pointed me to the code again and showed me that although the ECM limits one segment of memory to 255, it uses the correct segment for calculating the BPW. This is the code I am specifically referring ...

But, the code above does not limit the air flow since as the code continues it uses L006F (the real value) and not L006B (the limited value). My current line of thought focuses around the INV MAT code. Here's the specifics. Something is being limited to 255 here (LD7FA) ... and the INV MAT is used in the BPW calculation as per my previous post on this (https://www.thirdgen.org/techbb2/sho...hreadid=214156). FUTHERMORE, the INV MAT seems to be limited by L006B also. Can't quite figure it out. But I believe everything we need is staring right at us in the code I just posted.

 

Looking into this further today ...

It is documented how the Grams of Air per cylinder is calculated; namely,
GMS of Air / CYL = ((MAP - EGR part press) * CYL VOL)/((MAT + 233) * 128))

This is done in the ECM by ...
GMS of Air / CYL = ((MAP - EGR) * INV MAT) * VE

During this calc 2 values are stored. The real value is stored in L006F. A LIMITED value of 255 is stored in L006B if the real value is over L006F. I initially ruled out the limited value because the actual calculation of the Base Pulse Width uses the REAL value stored at L006F. Thus, the limited value is not utilized for calculating the BPW. However, it is used in serveral other calcs. I looked through all of these calcs and there is a routine at LD7DA with regard to calculating the INV MAT. As long as the flag for "Enable INV MAT lookup table" is checked, this routine is executed. It will begin by using the L006B limited airflow .... and also FORCE a max value of 255 later. I can't figure out exactly what is happening. But, it appears that limited values are being utilized when calculating and storing the INV MAT at L006D. I'm just not that good with assembly to see if this is definitely the problem or not. One test would simply be to burn a new chip with "Enable Inv. MAT Term Lookup Table" UNCHECKED. It is checked by default on the stock AUJP. This is a flag at hex 18 (mask 10). I don't know what unchecking this will do ... but, it should prevent this limited calc of the INV MAT and allow another calc of the INV MAT. Not sure how that would affect the car though. I wish I had my car for testing this today. I would test it immediately.

This is the most promising thing so far ... and I think we're on to something with this.

 

Trax, here is a thread where the operation of the inverse MAT code and tables is explained:

https://www.thirdgen.org/techbb2/sho...hreadid=265869


Just for info here is a thread about intake charge heating:

https://www.thirdgen.org/techbb2/sho...hreadid=184371

RBob.

 

Thanks Rob. I guess my uneducated hypothesis is null and void. I don't think this issue is going to get solved until somebody can run it on a bench. The bench is going to be necessary for this one. Somehow though ... I have a feeling that somebody already knows what is going on but just isn't posting about it. The 730 has been around way too long for this to have never been dealt with. The BPW is getting clamped (either indirectly before the BPW calculation or after the BPW calc). It's a fact. Just can't figure out why

Tim

 

I think there is something in the 165 ($6E) that does the same thing. I looked at a datalog of a friends car from his "original crappy tune" from a corvette owner that thinks the stock MAF is awesome and it was reporting pegged pulsewidths of 10ms from like 2600rpm and up. I'm going to adjust the injector size and see if we flatline the pulsewidth. He's got a 383 with 24lb injectors but a relatively small cam so I don't know if we'll actually SEE the flatline, but I'll try .